U.S. patent application number 10/352558 was filed with the patent office on 2004-07-29 for method of designing a thermal physical vapor deposition system.
This patent application is currently assigned to Eastman Kodak Company. Invention is credited to Freeman, Dennis R., Grace, Jeremy M., Klug, Justin H., Redden, Neil, Van Slyke, Steven A..
Application Number | 20040144321 10/352558 |
Document ID | / |
Family ID | 32655506 |
Filed Date | 2004-07-29 |
United States Patent
Application |
20040144321 |
Kind Code |
A1 |
Grace, Jeremy M. ; et
al. |
July 29, 2004 |
Method of designing a thermal physical vapor deposition system
Abstract
The need is met according to the present invention by providing
a method of designing a system for thermal vapor deposition that
includes a material to be deposited on a workpiece, an elongated
container for containing the material, a heater for heating the
material in the container to vaporize the material, the container
defining n apertures for emitting the vaporized material in an
elongated pattern in the elongated direction, that includes the
steps of: calculating the total source throughput Q per unit length
at a deposition rate of interest; calculating the internal pressure
P of the source required to produce Q for the total aperture
conductance C.sub.A of the source; modeling the system as a ladder
network of conductances, the elongated container having a container
conductance C.sub.B and conductances C.sub.b=nC.sub.B, between
apertures, and the apertures having a combined conductance 1 C A =
i = 1 n C ai , where C.sub.ai are individual aperture conductances;
and using the latter network model, designing the system to have a
desired pressure uniformity along the elongated direction of the
container.
Inventors: |
Grace, Jeremy M.; (Penfield,
NY) ; Freeman, Dennis R.; (Spencerport, NY) ;
Redden, Neil; (Sodus Point, NY) ; Klug, Justin
H.; (Rochester, NY) ; Van Slyke, Steven A.;
(Pittsford, NY) |
Correspondence
Address: |
Thomas H. Close
Patent Legal Staff
Eastman Kodak Company
343 State Street
Rochester
NY
14650-2201
US
|
Assignee: |
Eastman Kodak Company
|
Family ID: |
32655506 |
Appl. No.: |
10/352558 |
Filed: |
January 28, 2003 |
Current U.S.
Class: |
118/726 ;
118/722; 118/725 |
Current CPC
Class: |
C23C 14/243 20130101;
C23C 14/12 20130101; C23C 14/24 20130101 |
Class at
Publication: |
118/726 ;
118/722; 118/725 |
International
Class: |
C23C 016/00 |
Claims
What is claimed is:
1. A method of designing a system for thermal vapor deposition that
includes a material to be deposited on a workpiece, an elongated
container for containing the material, a heater for heating the
material in the container to vaporize the material, the container
defining n apertures for emitting the vaporized material in an
elongated pattern in the elongated direction, comprising the steps
of: a) calculating the total source throughput Q per unit length at
a deposition rate of interest; b) calculating the internal pressure
P of the source required to produce Q for the total aperture
conductance C.sub.A of the source; c) modeling the system as a
ladder network of conductances, the elongated container having a
container conductance C.sub.B and conductances C.sub.b=nC.sub.B,
between apertures, and the apertures having a combined conductance
29 C A = i = 1 n C ai , where C.sub.ai are individual aperture
conductances; and d) using the ladder network model, designing the
system to have a desired pressure uniformity along the elongated
direction of the container.
2. A system for thermal physical vapor deposition, comprising: a) a
material to be deposited on a workpiece; b) an elongated container
for containing the material, the container having a conductance
C.sub.B in the elongated direction; c) a heater for heating the
material in the container to vaporize the material to a partial
pressure Pm; d) the container defining one or more apertures for
emitting the vaporized material in an elongated pattern in the
elongated direction, the one or more apertures having a conductance
C.sub.A, wherein 30 C A C B 0.5 , and e) a means for providing
relative motion of the substrate and elongated container in a
direction substantially perpendicular to the elongated
direction.
3. The system claimed in claim 2, wherein the emission through the
apertures is by molecular flow and Pm.ltoreq.13 Pa.
4. The system claimed in claim 2, wherein the emission through the
apertures is by viscous or transition flow and Pm>13 Pa.
5. The system of claim 2, wherein the means for providing relative
motion between the elongated vapor deposition source and the
structure includes a lead screw adapted either to move the source
with respect to a fixedly disposed structure, or to move the
structure with respect to a fixedly disposed source.
6. The system claimed in claim 2, further comprising means for
introducing an inert gas into the container to decrease 31 C A C B
.
7. The system claimed in claim 6, wherein the inert gas is argon or
nitrogen.
8. The system claimed in claim 2 wherein the material is a
phosphorescent material, an electroluminescent material,
photoconducting, or luminescent by action of ionizing
radiation.
9. The system claimed in claim 2, wherein the materials are used to
make an OLED.
10. A method for coating a large-area substrate, comprising the
steps of: a) loading a material to be deposited on a workpiece into
an elongated container, the container having a conductance C.sub.B
in the elongated direction; b) heating the material in the
container to vaporize the material to a partial pressure Pm; c) the
container defining one or more apertures in an elongated pattern in
the elongated direction for emitting the vaporized material through
the apertures, the one or more apertures having a conductance CA;
wherein 32 C A C B 0.5 , and d) providing relative motion of the
substrate and elongated container in a direction substantially
perpendicular to the elongated direction.
11. The method claimed in claim 10, wherein the emission through
the apertures is by molecular flow and Pm<13 Pa.
12. The method claimed in claim 10, wherein the emission through
the apertures is by viscous or transition flow and Pm>13 Pa.
13. The method claimed in claim 10, wherein the container has a
cover having apertures, and further comprising a baffle between the
cover and the material to prevent vaporized material from passing
through the apertures in the cover without first engaging the walls
of the container.
14. The method claimed in claim 10, wherein 33 C A C B 0.1over a
desired range of operation.
15. The method claimed in claim 10, wherein the apertures have
varying size, shape or spacing between adjacent apertures, or
combinations thereof, selected to provide a substantially uniform
efflux of vaporized material along the elongated direction of the
container.
16. The method claimed in claim 10, wherein the method is used to
make an OLED.
17. The method claimed in claim 10, wherein the solid organic
material received in the container includes doped or undoped
organic hole-injecting material, doped or undoped organic
hole-transporting material, doped or undoped organic light-emitting
material, or doped or undoped organic electron-transporting
material.
18. The method claimed in claim 10, further comprising the step of
introducing an inert gas into the container to decrease 34 C A C
B
19. The method claimed in claim 18, wherein the inert gas is argon
or nitrogen.
20. The method claimed in claim 10 wherein the material is a
phosphorescent material, an electroluminescent material,
photoconducting, or luminescent by action of ionizing radiation.
Description
FIELD OF THE INVENTION
[0001] The present invention relates generally to vapor deposition
of an organic layer onto a structure which will form part of an
organic light-emitting device (OLED). More generally it relates to
coating layers used in producing luminescent devices, whether
electroluminescent, photoluminescent, luminescent by action of
ionizing radiation, or otherwise luminescent, as well as devices
made with photoconducting materials.
BACKGROUND OF THE INVENTION
[0002] An organic light-emitting device, also referred to as an
organic electroluminescent device, can be constructed by
sandwiching two or more organic layers between first and second
electrodes.
[0003] In a passive matrix organic light-emitting device (OLED) of
conventional construction, a plurality of laterally spaced
light-transmissive anodes, for example indium-tin-oxide (ITO)
anodes, are formed as first electrodes on a light-transmissive
substrate such as, for example, a glass substrate. Two or more
organic layers are then formed successively by vapor deposition of
respective organic materials from respective sources, within a
chamber held at reduced pressure, typically less than 10.sup.-3
torr (1.33.times.10.sup.-1 pascal). In addition to doped or undoped
organic light-emitting material, typical organic layers used in
making OLEDS are doped or undoped organic hole-injecting material,
doped or undoped organic hole-transporting material, and doped or
undoped organic electron-transporting material, where doping refers
to adding a minor constituent to enhance the electrical
performance, optical performance, stability, or life time of a
given material or device constructed thereof. A plurality of
laterally spaced cathodes is deposited as second electrodes over an
uppermost one of the organic layers. The cathodes are oriented at
an angle, typically at a right angle, with respect to the
anodes.
[0004] Applying an electrical potential (also referred to as a
drive voltage) operates such conventional passive matrix organic
light-emitting devices between appropriate columns (anodes) and,
sequentially, each row (cathode). When a cathode is biased
negatively with respect to an anode, light is emitted from a pixel
defined by an overlap area of the cathode and the anode, and
emitted light reaches an observer through the anode and the
substrate.
[0005] In an active matrix organic light-emitting device (OLED), an
array of anodes is provided as first electrodes by thin-film
transistors (TFTs) which are connected to a respective
light-transmissive portion. Two or more organic layers are formed
successively by vapor deposition in a manner substantially
equivalent to the construction of the aforementioned passive matrix
device. A common cathode is deposited as a second electrode over an
uppermost one of the organic layers. The construction and function
of an active matrix organic light-emitting device is described in
U.S. Pat. No. 5,550,066, the disclosure of which is herein
incorporated by reference.
[0006] Organic materials, thicknesses of vapor-deposited organic
layers, and layer configurations, useful in constructing an organic
light-emitting device, are described, for example, in U.S. Pat.
Nos. 4,356,429, 4,539,507, 4,720,432, and 4,769,292, the
disclosures of which are herein incorporated by reference.
[0007] Other kinds of imaging devices, such as imaging phosphors
for computed radiography and x-ray photoconductive devices for
digital radiography, depend on the ability to coat the active
materials uniformly over large areas. While the discussion below
pertains to organic light-emitting devices, it should be readily
apparent that the same invention can be applied to the deposition
of alkalihalide phosphors, amorphous semiconductors, and other
luminescent or photoactive layers, as well as a variety of other
materials used in devices based on such luminescent or photoactive
layers.
[0008] For sufficiently small substrates, a point source approach
can be implemented wherein the material to be deposited emanates
from a localized heated crucible and the substrate is placed
sufficiently far from the localized region of vaporization that the
coating is sufficiently uniform along the substrate. As substrate
size increases or working distance decreases, rotary or planetary
motion of the substrate relative to the localized source is often
required to produce the desired uniformity.
[0009] By elongating the vaporization source and providing for
translation of source and substrate relative to one another, the
desired uniformity can be attained at considerably smaller working
distances and thus considerably higher rates and better materials
utilization, if desired. Scaling of such a process to large areas
(i.e. substrates greater than 15 cm in at least one dimension) is
considerably easier than for point sources.
[0010] An elongated source for thermal physical vapor deposition of
organic layers onto a structure for making an organic
light-emitting device has been disclosed by Robert G. Spahn in
commonly assigned U.S. Pat. No. 6,237,529, issued May 29, 2001. The
source disclosed by Spahn includes a housing, which defines an
enclosure for receiving solid organic material, which can be
vaporized. The housing is further defined by a top plate which
defines a vapor efflux slit-aperture for permitting vaporized
organic materials to pass through the slit onto a surface of a
structure. The housing defining the enclosure is connected to the
top plate. The source disclosed by Spahn further includes a
conductive baffle member attached to the top plate. This baffle
member provides line-of-sight covering of the slit in the top plate
so that vaporized organic material can pass around the baffle
member and through the slit onto the substrate or structure while
particles of organic materials are prevented from passing through
the slit by the baffle member when an electrical potential is
applied to the housing to cause heat to be applied to the solid
organic material in the enclosure causing the solid organic
material to vaporize.
[0011] In using the thermal physical vapor deposition source
disclosed by Spahn to form an organic layer of a selected organic
material on a substrate or structure, it has been found that the
vapor efflux slit-aperture causes nonuniform vapor flux of organic
material vapor to emanate along a length dimension of the slit.
While the technical or physical aspects of source design related to
this nonuniformity of vapor flux are not fully understood at
present, it appears that opposing edges of the slit-aperture, i.e.
edges opposed in a width direction of the slit, sag or rise
nonuniformly over a central portion of the slit when the source is
heated to cause vaporization of solid organic material. This is a
particular problem when a width dimension of the slit is reduced,
for example, to a width dimension less than 0.5 millimeter (mm).
Such spatially nonuniform orientation of opposing slit edges can be
thought of as a deviation of planarity of opposing edges which, in
turn, can promote a greater fraction of vaporized organic material
to exit the vapor deposition source through a central portion of
the slit, with a correspondingly lower fraction of vaporized
organic material exiting the source through remaining portions of
the slit along its length dimension. Such nonuniform vapor flux,
directed at a substrate or structure, will cause the formation of
an organic layer thereon which will have a nonuniform layer
thickness in correspondence with the nonuniform vapor flux.
[0012] In addition, any nonuniformities in heat generation from the
heater or heat absorption by the material to be deposited or
distribution of said material within the source can give rise to
nonuniformity in deposition along the length of the source. Yet
another potential source of nonuniformity is unintended leaks in
the source enclosure other than the apertures used to deliver the
material vapor. If such leaks exist at the ends of the source, the
flow of vapor from center to end of the source can cause pressure
gradients within the source, thereby causing nonuniformity in the
resultant deposition.
[0013] Forrest et al. (U.S. Pat. No. 6,337,102 B1) disclose a
method for vaporizing organic materials and organic precursors and
delivering them to a reactor vessel wherein the substrate is
situated and delivery of the vapors generated from solids or
liquids is accomplished by use of carrier gases. In one embodiment
of their invention, Forrest et al. locate the substrates within a
suitably large reactor vessel, and the vapors carried thereto mix
and react or condense on the substrate. Another embodiment of their
invention is directed towards applications involving coating of
large area substrates and putting several such deposition processes
in serial fashion with one another. For this embodiment, Forrest et
al. disclose the use of a gas curtain fed by a gas manifold
(defined in the disclosure as "hollow tubes having a line of
holes") in order to form a continuous line of depositing material
perpendicular to the direction of substrate travel.
[0014] The approach to vapor delivery as disclosed by Forrest et
al. can be characterized as "remote vaporization" wherein a
material is converted to vapor in an apparatus external to the
deposition zone and more likely external to the deposition chamber.
Organic vapors, alone or in combination with carrier gases are
conveyed into the deposition chamber and ultimately to the
substrate surface. Great care must be taken using this approach to
avoid unwanted condensation in the delivery lines by use of
appropriate heating methods. This problem becomes even more
critical when contemplating the use of inorganic materials that
vaporize to the desired extent at substantially higher
temperatures. Furthermore, the delivery of the vaporized material
for coating large areas uniformly requires the use of gas
manifolds. No mention of the requirements for such a gas manifold
is made by Forrest et al.
[0015] As can be appreciated from the disclosure of Forrest et al.,
one skilled in the art would expect to have difficulty providing
uniform films from an elongated source in which the material is
vaporized along the length of a deposition source within the
deposition zone. There is a need therefore for an improved method
of designing a thermal physical vapor deposition system in which
the material is vaporized along the length of a deposition source
within the deposition zone that avoids the problems noted
above.
SUMMARY OF THE INVENTION
[0016] The need is met according to the present invention by
providing a method of designing a system for thermal vapor
deposition that includes a material to be deposited on a workpiece,
an elongated container for containing the material, a heater for
heating the material in the container to vaporize the material, the
container defining n apertures for emitting the vaporized material
in an elongated pattern in the elongated direction, that includes
the steps of: calculating the total source throughput Q per unit
length at a deposition rate of interest; calculating the internal
pressure P of the source required to produce Q for the total
aperture conductance C.sub.A of the source; modeling the system as
a ladder network of conductances, the elongated container having a
container conductance C.sub.B and conductances C.sub.b=nC.sub.B,
between apertures, and the apertures having a combined conductance
2 C A = i = 1 n C ai ,
[0017] where C.sub.ai are individual aperture conductances; and
using the ladder network model, designing the system to have a
desired pressure uniformity along the elongated direction of the
container.
[0018] A novel system and method resulting from the design method
includes a material to be deposited on a workpiece; an elongated
container for containing the material, the container having a
conductance C.sub.B in the elongated direction; a heater for
heating the material in the container to vaporize the material to a
partial pressure Pm; the container defining one or more apertures
for emitting the vaporized material in an elongated pattern in the
elongated direction, the one or more apertures having a conductance
C.sub.A; and wherein 3 C A C B 0.5 .
ADVANTAGES
[0019] The present invention demonstrates an unanticipated way in
which the design of the gas manifold depends on the operating
pressure and provides a means to overcome the problems one skilled
in the art would reasonably expect to encounter. An advantage of
the present invention is that the design based on the conductance
ratio ensures tolerance to nonuniformities in vaporization of
material in the container, these nonuniformities arising from a
variety of sources, including, non-uniform heating of the heater,
nonuniform distribution of the material in the container, and
non-uniform application of radiant heat from the heater surface to
the surface of the material in the container.
[0020] Another advantage of the present invention is that the
conductance criterion (i.e. C.sub.A/C.sub.B<a specified value)
ensures tolerance to significant vapor flow along the length of the
source. Such vapor flow can arise from condensation of material at
the ends of the source or unintentional leaks at either or both
ends of the source.
[0021] Another advantage of the present invention is that the vapor
deposition source can be made to operate over a wide range of
deposition rate from the desired minimum deposition rate for the
material to be deposited to substantially higher rates, or a
suitably inert gas may be used to provide internal pressure
sufficient to maintain adequate gas conductance along the vapor
deposition source length and to allow operation from the highest
intended deposition rates down to arbitrarily low rates.
BRIEF DESCRIPTION OF THE DRAWINGS
[0022] FIG. 1 is a schematic view of an elongated vapor deposition
source having a rectangular cross section that can be designed
according to the present invention.
[0023] FIG. 2 is a schematic view of an elongated vapor deposition
source having a circular cross section that can be designed
according to the present invention.
[0024] FIG. 3 is a schematic representation of the conductances
governing the emission of material from the vapor deposition
source.
[0025] FIG. 4 is a circuit model of the worst-case scenario for
deposition using a source as described in FIGS. 1, 2 and 3.
[0026] FIG. 5 is a schematic cross sectional view of an elongated
vapor deposition source being used to coat a large substrate.
[0027] FIG. 6 is a schematic sectional view of a vapor deposition
station dedicated to forming vapor-deposited layers and showing a
structure being moved by a lead screw with respect to a stationary
vapor deposition source to provide a uniformly vapor-deposited
layer over the substrate, in accordance with an aspect of the
present invention.
[0028] FIG. 7 is a graph of vapor pressure as a function of
1/Temperature for AlQ, as measured by thermal gravimetric
analysis.
[0029] FIG. 8 is a graph of the pressure nonuniformity along the
source length as a function of deposition rate for examples
1-5.
[0030] FIG. 9 is a graph of the pressure nonuniformity inferred
from observed rate nonuniformity as a function of the experimental
deposition rate at an aperture-substrate spacing of 10 cm and for
two different degrees of internal mass flow arising from
condensation at the ends of the source.
[0031] FIG. 10 is a graph of the calculated pressure drop along the
source length as a function of cross sectional area of the source
for examples 1-5.
[0032] FIG. 11 is a graph of the calculated pressure nonuniformity
along the source length as a function of the conductance ratio for
examples 1-5. Data are shown for no leak at the far end and a leak
equal to the total aperture conductance.
[0033] FIG. 12 is a flow chart for the calculation of conductances
and pressure distribution in the deposition source.
DETAILED DESCRIPTION OF THE INVENTION
[0034] The present invention provides a method of designing a vapor
deposition system that includes a heated container in which a
material is vaporized (whether it be a heated crucible, a heated
bubbler, or some other form of heated container) and a gas delivery
manifold defined by one or more apertures in the container.
[0035] Turning to FIGS. 1 and 2, respectively perspective views of
an elongated vapor deposition source having a rectangular cross
section and having a circular cross section are shown. The source
body or container 1,10 is filled with material 2, 11. A cover or
lid 3, 12 seals the source body. In FIGS. 1 and 2, apertures 4, 13
are located in a row along the central axis of the cover 3, 12.
While the apertures are shown in the top surface, which is used to
seal the container, the apertures need not be in the top surface,
and the sealing surface (i.e. cover or lid) need not be the top
surface of the container.
[0036] The apertures can be in any elongated pattern, including
multiple rows, staggered or aligned, and they can be any shape,
including circular, rectangular, elliptical, ovular, or square.
Furthermore, the cover or lid can either seal to the source body
directly or it may cover an opening such as the large slit 16
(shown by dotted lines) in FIG. 2. Clamps or wires of a variety of
kinds, such as tension bands, screwed vise clamps, internal flanges
with bolts and tapped holes or external flanges with through holes
and nuts, etc. can be used to affix the cover or lid in contact
with the source body.
[0037] A baffle 6, 15 prevents particulate matter from being
ejected towards the apertures and ensures that vapor exiting the
apertures has collided with the internal surfaces of the container
so the material flow is well established. The cover or lid can be
electrically conductive and the body or container electrically
insulating, or vice versa, or both can be insulating or both can be
conductive. End contacts 5, 14 can be used for mechanical fixture
and for making electrical contact to the top surface, when the top
surface is used as a resistive heater. Alternatively, electrical
feedthroughs (not shown) can be placed through a long surface of
the container to make electrical contact to either the baffle 6, 15
or an additional heater internal to the source body (not shown).
The surface with apertures can be heated separately from the baffle
or with the baffle (or internal heating element) in serial or
parallel fashion. Alternatively, the source can be radiantly heated
by external means (not shown).
[0038] FIG. 3 illustrates how the source body and apertures can be
represented as respective resistances R.sub.B and R.sub.a. The
source body is represented as one resistance R.sub.B and the
apertures are represented as a plurality of parallel connected
resistors R.sub.a. The internal pressure P of the vapor in the
source is analogous to the voltage V along one side of the parallel
resistors R.sub.a.
[0039] In a worst case, shown in FIG. 4, vapor is generated largely
at one end of the source, corresponding to a voltage V being
applied at one end of the parallel array of resistors R.sub.a.
Furthermore, each section of the source body has some resistance
R.sub.b between apertures. The result is that the pressure drops
along the length of the source, thereby causing non uniform
deposition.
[0040] Now turning to FIG. 5, a material 30 to be deposited on
substrate 40 is loaded into the source body 31. The source body can
be made from a thermally insulating material or a thermally
conductive material. It can also be electrically insulating or
conducting. Furthermore, the source body can contain within it a
crucible for holding material, said crucible being made from a
thermally insulating material or a thermally conductive material.
The terms "substrate" and "workpiece" as used herein denote a
support on which a luminescent or otherwise photoactive device or
array thereof is constructed. The term "structure" is used to
describe the substrate once it has received a portion of a vapor
deposited layer.
[0041] An aperture plate 32 serving as the emitting surface can be
affixed to the source body as a separate piece and sealed thereto
by use of fasteners and a high-temperature gasket material (for
example, Grafoil.TM.), by welding, or it can be an integral part of
the source body itself. In the former case, the material can be
loaded by removal of the aperture plate and any associated
fixtures. In the latter cases, material must be loaded through an
opening elsewhere that can later be sealed (for example, at the
ends of the source body or an opening cut in the side of the source
body). An array of apertures 33 located on the aperture plate 32 is
employed to distribute the vaporized material 30 uniformly along
the length of the source onto substrate 40. The spacing of the
apertures can be tailored near the ends of the source in order to
compensate for edge effects that generally cause lower deposited
thickness at the edges of the substrate. The source-substrate
distance 35 is defined as the vertical distance from the plane of
the upper surface of the aperture plate 32 to the plane of the
lower surface of the substrate 40.
[0042] A baffle 34 as disclosed by Spahn is used to prevent direct
line of sight between vaporizing material 30 and the apertures 33.
The baffle may also serve as the active heating element, or
separate heating elements can be used, either internal to the
source or by external radiative means. For example, the aperture
plate can be electrically isolated and used as a resistive heating
element. Alternatively, the baffle can be electrically isolated and
used as a resistive heating element, or an additional heating
element can be inserted between the baffle and aperture plate, or
below the baffle. As yet another example, an external heater may
also be used to provide a substantially uniform temperature within
the interior of the elongated vapor deposition source.
[0043] Regardless of heating method, it may be advantageous to have
one or more radiation shields 50 to ensure that the temperature of
the exit surfaces of the apertures remains sufficiently high to
prevent condensation thereon and to limit the temperature excursion
of the substrate during the deposition process. Furthermore, such
radiation shields can be used to reduce the input power required
for achieving the desired vapor emission rate from the source.
[0044] To coat large areas, the source is made sufficiently long
and then is translated relative to the substrate in the direction
perpendicular to its long axis. For example, the source can be
placed on standoffs 42 and affixed to a carriage 41, which can be
motorized and passed underneath the substrate 40. Alternatively,
the substrate fixture can be moved relative to a fixed source or
both substrate and source can be moved to achieve the desired
relative translation.
[0045] A key requirement for uniform deposition from an elongated
source (such as depicted in FIGS. 1-5) is that the pressure
distribution along the length of the source is repeatable. In
general, if the pressure distribution is known, the apertures can
be sized accordingly such that the product C.sub.iP.sub.i, i.e. the
throughput from each aperture, is maintained constant along most of
the source length and adjusted appropriately near the edges in
order to compensate the edge effects associated with the finite
length of the source (here, C.sub.i is the conductance of an
aperture and P.sub.i is the pressure in the source at the aperture
location.). The Conductances C.sub.i are the reciprocals of the
resistances employed in the model described above. A special case
of a repeatable pressure distribution is one that is nearly
constant along the length of the source and, furthermore, is
insensitive to nonuniform vaporization (from temperature
non-uniformities), as well as flow-induced pressure gradients from
vaporized material exiting the source via the apertures or leaks in
the source other than the apertures.
[0046] To provide a substantially uniform and robust pressure
within the source body, the relative conductance of the apertures
and the body must be taken into account. By calculating
conductances and considering a worst-case scenario for flow of the
material vapor through the source body, a conductance
criterion--i.e., that the ratio of the total conductance of all
apertures to the conductance of the source body be below some
value--emerges. The basis for the conductance criterion is
explained below.
[0047] The mean free path in an ideal gas depends on temperature,
pressure, and the molecular size (see J. M. Lafferty, ed,
Foundations of Vacuum Science and Technology, Wiley & Sons, New
York, 1998, p.8)
L=k.sub.BT/({square root}{square root over
(2)}.multidot.P.pi..multidot..d- elta..sup.2)
[0048] where L is in m, k.sub.B is the Boltzmann constant, P is the
pressure in Pa, and .delta. is the molecular diameter in m. A
convenient form of this expression is 4 L = 5.3 10 - 6 ( T / 273 )
( / 4 ) 2 ( P / 10 5 ) ,
[0049] where L is in cm, T is in K, .delta. is in .ANG., and P is
in Pa. This form for L is used to estimate the Knudsen number
K.sub.n=L/d,
[0050] where d is a characteristic dimension in the system.
[0051] During the vapor deposition of organic materials using an
elongated source such as described in this invention, the vaporized
organic material is emitted from the emitting surface through a
series of apertures. In general, these apertures are small enough
in at least one dimension such that Knudsen number (i.e., the ratio
of mean free path to said small dimension) exceeds the value 1 and
the aperture is said to be in molecular flow.
[0052] In molecular flow, the conductance of an orifice (or
aperture) having zero length is given by (see Lafferty, ibid p 86):
5 C 0 = A R 0 T 2 M m ,
[0053] where A is the orifice area in m.sup.2, R.sub.0 is the
universal gas constant, T is the absolute temperature, and M.sub.m
is the molar mass in kg. This expression is closely related to the
emission rate discussed below. For air at 22 C, C.sub.0=11.6 A in
1/s and A is in cm.sup.2. For other gases at different
temperatures, C.sub.0 can be expressed as 6 C 0 ( l / s ) = 11.6 A
( cm 2 ) T ( K ) / 293.16 M m ( g ) / 28.8 .
[0054] As the orifice size decreases, the thickness of the orifice
walls becomes increasingly important. Eventually, the orifice
becomes more like a duct. The molecular flow conductance of a duct
is similar to that of the zero-length orifice, but is modified by
the transmission probability: 7 C a = A R 0 T 2 M m .
[0055] Here, C.sub.a is the conductance of an aperture of non-zero
length. Transmission probabilities .alpha. for various geometries
have been derived by various means. For certain simple geometries,
analytical solutions can be found. For more complex geometries,
Monte Carlo techniques are used. Values of .alpha. are tabulated
for a variety of geometries, including short ducts of circular
cross section and short ducts of rectangular cross section (see
Lafferty, ibid, p. 91 and O'Hanlon, J. F, A User's Guide to Vacuum
Technology, 2.sup.nd ed, John Wiley & Sons, New York, 1989, pp.
36-37).
[0056] The throughput Q for each aperture is given by:
Q=C.sub.aP,
[0057] where C.sub.a is the conductance defined above, and P is the
pressure drop across the orifice. Thus, uniform emission from an
elongated source with a plurality of apertures requires that the
pressure drop P across each aperture be as uniform as possible. As
the organic vapor travels through the vapor source body to the
apertures, the conductance to gas flow along the source body must
be sufficient to maintain nearly constant pressure within the
deposition source. By allowing the vapor to move freely along the
source length, the adverse effects of localized non-uniform
vaporization, as well as localized leaks in the source
construction, will be minimized, and uniform deposition can be
achieved.
[0058] The conductance C.sub.B along the vapor deposition source
can be calculated by determining the molecular flow conductance
along the source body, the viscous flow conductance along the
source body, and the Knudsen number. As discussed in O'Hanlon (ibid
pp. 26-27), depending on the Knudsen number, the conductance is
either taken to be the molecular flow value (K.sub.n>1) or the
viscous flow value (K.sub.n<0.01), or a linear combination of
the two values (0.01<K.sub.n<1).
[0059] For a long cylindrical tube of diameter d and length l in
molecular flow, the transmission probability can be calculated
analytically and the molecular flow conductance can be expressed
as: 8 C ( l / s ) = 12.1 d 3 ( cm 3 ) l ( cm ) T ( K ) / 293.16 M m
( g ) / 28.8 .
[0060] For a long rectangular duct with minor width b, major width
a, and length l (all in cm), the molecular flow conductance can be
expressed as: 9 C ( l / s ) = 11.6 ba [ 16 3 3 / 2 b 1 ln ( 4 a b +
3 b 4 a ) ] T ( K ) / 293.16 M m ( g ) / 28.8
[0061] (Note that the convention adopted here for defining a and b
is that used by O'Hanlon (ibid, page 35) and is the reverse of that
used by Lafferty (op cit., page 90).
[0062] In viscous flow, the conductance of a duct increases
linearly with the average pressure in the duct and inversely as the
viscosity of the gas. For a long circular tube, the viscous flow
conductance can be expressed as: 10 C ( l / s ) = 1.38 d 4 ( cm 4 )
l ( cm ) < P > ( ( ) / 3.74 ) 2 ( T ( K ) / 273.16 ) ( M m (
g ) / 28.8 ) ,
[0063] where .delta. is the molecular diameter, which has been
taken as 3.74 .ANG. for air (Lafferty, ibid, p. 9, 34), and it has
been assumed that the viscosity of the gas obeys the following
relationship (see O'Hanlon, op cit. p. 17) 11 M m k B T 2 .
[0064] For a long rectangular duct in viscous flow, the conductance
can be expressed as (see Lafferty op cit. p. 111 and O'Hanlon op
cit. p. 31): 12 C ( l / s ) = 4.6 [ a 3 b 3 / ( a 2 + b 2 + 0.371 a
b ) ] < P > l ( ( ) / 3.74 ) 2 ( T ( K ) / 293.16 ) ( M m ( g
) / 28.8 )
[0065] where a, b, and l are in cm and are as defined previously
for rectangular ducts, and <P> is the average pressure in the
duct in Pa.
[0066] The method disclosed here relies upon comparing the aperture
and boat conductances and is not specific to which conduction
regime is in effect. In the examples disclosed herein the
geometries, molecular properties, temperatures, and pressure ranges
are such that the orifices are short ducts in molecular flow, while
the source body is treated as a long duct in molecular, transition,
or viscous flow (depending on cross sectional dimensions, pressure,
molecular size, and temperature). Furthermore, in the case of
materials such as aluminum tris-quinolate (AlQ), the corrections
for T, M.sub.m, and .delta. are significant--the molecular flow
conductance is reduced considerably (relative to air at 20 C) by
the high molecular mass (and hence slower thermal velocity relative
to air), while the viscous flow is slightly increased by the net
decrease in viscosity, even with the increase in molecular mass and
temperature (see the viscosity relationship above).
[0067] In order to determine the viscous flow conductance, the
average pressure in the vapor deposition source must be known. The
operating pressure of the linear source can be estimated from known
deposition rates at a known distance from the source. A simple
estimate can be made treating the source as a line source. Applying
Gauss' law and making use of the cylindrical symmetry of the
situation, one finds that the deposition rate r directly above the
center of the source depends inversely on the distance d from
source to substrate: 13 r = Q l d .
[0068] Here, Q is the mass per unit time emitted from the source of
length l, and it has been assumed that the source emits into the
semi cylinder of radius d. The factor of .pi.d arises from the
integral of the isotropic flux density over the surface of the
semi-cylinder.
[0069] This estimate can be improved by taking into account the
emission profile of the source (see Lafferty op cit. pp. 85-86).
For thin apertures, the emission profile follows a cos(.theta.)
distribution (the magnitude of the flux density emanating from the
source at an angle .theta. with respect to the normal to the plane
of the source is proportional to cos(.theta.)). For emission from
deeper structures (i.e., thick apertures or tubes) the flux density
is significantly more anisotropic--the larger the ratio of the
length of the opening to the lateral dimension, the more beam-like
the emission (see Lafferty ibid, p. 86).
[0070] At sufficiently low operating pressures within the source,
the flow through the apertures can be considered to be molecular
flow, and the beaming effects (i.e. flow that appears more like a
beam of particles as opposed to a cloud) are approximated by
introducing a factor of cos .sup.p(.theta.), where p is a power
empirically determined to account for the resultant deposition
profile. At higher operating pressures, the flow through the
apertures may have a significant contribution from viscous flow. In
this case as well, the ratio of length of opening to the lateral
dimension influences the degree of beaming.
[0071] With a higher degree of beaming, the source material is used
more efficiently to produce a coating directly above the source
(the contribution to the integral of the flux density over the
surface of the semi-cylinder is significantly greater for small
values of .theta. than for larger values of .theta.). From the
measured deposition rate directly over the center of the elongated
source at a specified distance from the source to the substrate,
the estimated flux density emanating from the source is given as
follows: 14 q / l = r 10 - 8 1 d , where I = - 2 2 cos p ( )
[0072] Here, q/l is in g/cm/s, r is in (.ANG./s), d is the distance
from the center of the source to the substrate in cm, and the mass
density p of the coating is assumed to be 1 g/cm.sup.3. This
expression is valid when d is large compared to the lateral
distance over which the deposition rate is sampled. For
measurements of r over a flat substrate with edge
dimension.times.cm, the expression for I is: 15 I = - 2 2 cos p ( )
1 x - x 2 x 2 cos p + 2 ( ( x ' ) ) x ' ,
[0073] where .theta.(x) is the angle between the vector from the
source axis to a point at x on the substrate and the normal to the
source axis (pointing to x=0). The extra powers of cos(.theta.) in
the integral in the denominator arise from taking the dot product
of the flux density vector with the substrate surface normal and
from the change in distance to the source as a function of x.
[0074] The pressure in a source emitting material at rate q is
calculated assuming that the source material is at equilibrium
vapor pressure: 16 P = q N A A o M w 2 k B T m .
[0075] Here, q is the material emission rate in g/s, A.sub.o is the
total aperture area in m.sup.2, T is the temperature in Kelvin of
the vapor of material being deposited, M.sub.w is the molecular
weight of the material being deposited (in amu), m is the mass in
kg of a molecule of the material being deposited, N.sub.A is
Avogadro's number, and k.sub.B is the Boltzmann constant. The
pressure thus calculated is in Pascal. This expression is valid for
orifices having zero wall thickness (i.e., in the limit that the
orifice/heater plate is much thinner than the orifice
dimension).
[0076] In this calculation, the rate of emission of material from
the source is taken from kinetic theory (Lafferty op cit. pp.
18-21) and depends on the density n and thermal velocity v.sub.T of
the source vapor: 17 Emission rate ( molec . / s ) = ( nv T 4 ) A
,
[0077] where A is the orifice area, and the thermal velocity is
taken from the kinetic theory of gases:
v.sub.T={square root}{square root over (8k.sub.BT/.pi.m)}.
[0078] For non-zero aperture thickness, the expression above must
be divided by the transmission probability .alpha. for an
individual orifice, which is now treated as a short duct in
molecular flow: 18 P = q N A A o M w 2 k B T m .
[0079] For operation at higher pressures, one must consider
transition and viscous flow through the apertures. While the
example presented here is for molecular flow through the apertures,
it should be readily apparent to those skilled in the art that the
emission from the source and the conductance of the apertures can
be modified for the cases of transition and viscous flow, and thus
the respective determination of operating pressure and aperture
conductance can be made for the cases where the apertures are not
in molecular flow.
[0080] A worst-case scenario for non-uniform vaporization is if the
vapor enters the source from only one end (similar to loading the
solid material only on one end of the source). The resultant
pressure drop can then be calculated as a function of source
geometry and deposition rate.
[0081] As described above, given the deposition rate, the distance
from source to substrate, and the material being deposited, a
source emission rate is calculated. The emission rate and the
orifice geometry (hence, the transmission probability) are then
used to calculate the source pressure. The vaporization curve of
the material (see. Lafferty, ibid., pp. 22-25 for a discussion of
vapor pressure) is then used to estimate the gas temperature within
the source. The source pressure is then adjusted using the new
value for temperature. Generally this process converges in only a
few iterations, as vaporization curves have very strong temperature
dependences, and only a small change in temperature is required to
produce a rather large change in pressure.
[0082] From the calculated source pressure and temperature, the
conductance of the source body is then calculated by treating it as
a long hollow tube. (The internal baffle is neglected in this
calculation, but the upper surface of the material loaded into the
source is considered to be one of the walls of the source body.
Conductances are calculated for both the full source and the empty
source.) A simple conductance ratio is calculated by taking the sum
of the conductances of all the apertures, i.e. N.times.C.sub.a,
where N is the number of apertures, and dividing that sum by the
conductance along the source body. For a variety of deposition
rates and source geometries, the effect on the simple conductance
ratio can be assessed.
[0083] As the elongated vapor deposition source consists of an
enclosure with exit apertures, it lends itself to treatment as a
gas distribution manifold. As a limiting case, one can consider the
material to be deposited as entering as a vapor from one end of the
source, with the other end of the source closed off. Coating
uniformity thus depends critically on pressure uniformity along the
length of the source. This scenario is a worst-case scenario and
would correspond to an extremely non-uniform vaporization of the
source material, either from localized intense heating, or from
non-uniform distribution of material (e.g., all source material
loaded at one end of the boat). Hence, the ability to maintain
constant pressure along the source body in the presence of
localized heating or non-uniform distribution of material
translates to the ability of a gas delivery manifold to operate at
constant pressure along its length.
[0084] The elongated vapor deposition source can be considered a
series of segments, each with an exit aperture and a short tube
length. The short lengths of tube are considered to be resistances
(resistance=1/conductanc- e) in series, which form one side of a
ladder network. The exit orifices are parallel resistances to
ground (vacuum), which form the rungs of the ladder (see FIGS. 3
and 4). Because of the considerable orifice effects for short ducts
(see O'Hanlon op cit. pp. 45-47), the series resistances are taken
to be the net resistance of the source body divided by the number
of segments, rather than the resistance that would be calculated
from the conductance of an individual segment, thus
C.sub.b=N.times.C.sub.B. (Here, the number of segments N is the
number of exit apertures, assumed to be equally spaced for the
purposes of calculating the conductance ratio C.sub.A/C.sub.B.)
[0085] By an iterative procedure, the net conductance of the entire
ladder network is calculated. (The first orifice is added in series
with the first segment body. That pair is added in parallel with
the next orifice. The next segment body is added in series with the
previous net conductance. The next orifice is added in parallel
with the new net conductance, and so on.) The input mass flow to
the manifold is calculated from the required average deposition
rate. By a similar iterative procedure, the integrated flow through
the orifices and the pressure (in the source body segments) are
calculated along the length of the network. The inlet pressure is
then adjusted so that the average pressure in the source body
corresponds to that required to obtain the specified deposition
rate. The pressure difference from the inlet to the opposite end of
the source is then divided by the average pressure to obtain the
pressure nonuniformity along the source body.
[0086] In order to assess sensitivity to leaks, the ladder network
was modified to include a leak conductance on the end opposite that
from which the vapor enters. The net pressure nonuniformity is then
calculated by taking the pressure difference from inlet to leak and
dividing by the average pressure. This value includes both the
effects of the flow through the apertures and the flow induced by
the leak.
[0087] Large-area coating using the present invention can be
implemented for coating of layers used in making OLEDs and other
optically active devices. As an example, FIG. 6 shows a schematic
sectional view of a vapor deposition system of the present
invention. The vapor deposition station 100 has a housing 101,
which defines a chamber 102. A substrate or structure 110 is
supported in a holder and/or in a mask frame 111 within the chamber
102 which is at reduced pressure, typically at a pressure lower
than 0.13 Pa.
[0088] The thermal physical vapor deposition source 114 is
supported by a thermally and electrically insulating source support
115. Electrical leads 116 and 117 are schematically shown directed
toward the source from respective power feedthroughs 120 and 121
disposed in the housing 101.
[0089] In FIG. 6, relative motion between the substrate or
structure 110 and the vapor deposition source, during vapor
deposition of material 130 in a deposition zone 131 of vaporized
material, is provided by moving or translating the substrate or
structure 110 with respect to the source. The vapor deposition
source, i.e. the plurality of apertures 132, has a spacing D from
the substrate or structure 110.
[0090] In an intermediate vapor deposition position "II", the
substrate or structure 110, the holder and/or mask frame 111, a
glide shoe 140, and a lead screw follower 142 are shown in
solid-outline sectional view. These source elements are depicted in
dotted and dashed outlines in a left position "I" of the holder
111, and in right position "III" of the holder, these positions
being respectively the start and finish of a rightward motion of
the holder or the finish and start of a leftward motion of the
holder.
[0091] Leftward and rightward motion is effected by a lead screw
145 which engages the lead screw follower 142. The follower 142 is
attached to the glide shoe 140, which, in turn, supports the holder
and/or mask frame 111. The glide shoe 140 glides along a glide rail
147. The lead screw shaft 150 extends through the housing 130 and
shaft seal 149 to a motor 151.
[0092] The motor 150 provides for leftward or rightward motion or
idling (no motion) in a designated position in which a substrate or
structure 110 with a completed organic layer is removed from the
holder and/or mask frame 111 and a new substrate or structure is
positioned in the holder.
[0093] In FIG. 6, the substrate or structure 110 is moved with
respect to a stationary elongated vapor deposition source in a
direction substantially perpendicular to the elongated direction of
the source. Alternatively, relative motion between the substrate or
structure 110 and the elongated vapor deposition source can be
provided by moving the source with respect to a stationary
substrate or structure by a lead screw which engages a movable
carriage or other movable transport means on which the elongated
vapor deposition source can be positioned.
[0094] The elongated thermal physical vapor deposition source of
the present invention can also be effectively used to form a
uniform layer of one or more organic host materials and one or more
organic dopant materials by vapor deposition from one elongated
source having a plurality of vapor efflux apertures. The host
material(s) and the dopant material(s) are received in an elongated
container 2,11,31 in the form of powders, flakes, or particles, or
in the form of agglomerated pellets.
[0095] The invention and its advantages are further illustrated by
the following specific examples.
EXAMPLE 1
[0096] A 60 cm long source is covered with an aperture plate and
heater assembly having 60 apertures with their centers spaced 0.9
cm apart, starting 2.55 cm from the end of source. The apertures
are rectangular slots 0.0125 cm wide.times.0.5 cm long. The source
body is rectangular in cross section, having width b and depth a,
where the depth a denotes the distance from the aperture plate to
the surface of the material loaded in the source. The deposition
rate is 50 A/s at an aperture-substrate distance of 12 cm, and the
material being vaporized is aluminum tris-quinolate (AlQ) (having a
molecular mass of 459, a molecular diameter of 1 nm, and a vapor
pressure curve as given in FIG. 7). Values for a and b are shown in
Table I. The conductance ratio is calculated for the required
operating pressure to produce the specified rate and is shown in
Table I. The pressure nonuniformity is calculated as described
above. Results are shown in Table I for the case where there are no
leaks (other than the apertures) and for the case where there is a
leak at one end of the source, said leak having a conductance equal
to the total conductance of all the apertures.
1TABLE I Source Width, b (cm) Source depth, a (cm) Source Length
(cm) 19 Conductance Ratio : C apertures C sourcebody ( C A / C B )
20 Pressure nonuniformity 100 ( P max - P min ) ( P max + P min ) /
2 Pressure nonuniformity with leak C.sub.leak = C.sub.A 0.611 2.54
60 2.01 87.4 154.3 0.8 2.54 60 1.03 48 102.1 1 2.54 60 0.59 28.6
68.4 2 2.54 60 0.12 6 16.9 2.54 2.54 60 0.078 3.9 11.3 2.54 3 60
0.058 2.9 8.4 2.54 3.5 60 0.044 2.2 6.4 2.54 4 60 0.035 1.8 5.2
EXAMPLE 2
[0097] A 60 cm long source is covered with an aperture plate and
heater assembly having 60 apertures with their centers spaced 0.9
cm apart, starting 2.55 cm from the end of source. The apertures
are rectangular slots 0.0125 cm wide.times.0.5 cm long. The source
body is rectangular in cross section, having width b and depth a,
where the depth a denotes the distance from the aperture plate to
the surface of the material loaded in the source. The deposition
rate is 10 A/s at an aperture-substrate distance of 12 cm, and the
material being vaporized is aluminum tris-quinolate (AlQ) (having a
molecular mass of 459, a molecular diameter of 1 nm, and a vapor
pressure curve as given in FIG. 7). Values for a and b are shown in
Table II. The conductance ratio is calculated for the required
operating pressure to produce the specified rate and is shown in
Table II. The pressure nonuniformity is calculated as described
above. Results are shown in Table II for the case where there are
no leaks (other than the apertures) and for the case where there is
a leak at one end of the source, said leak having a conductance
equal to the total conductance of all the apertures.
2TABLE II Source Width, b (cm) Source depth, a (cm) Source Length
(cm) 21 Conductance Ratio : C apertures C sourcebody ( C A / C B )
22 Pressure nonuniformity 100 ( P max - P min ) ( P max + P min ) /
2 Pressure nonuniformity with leak C.sub.leak = C.sub.A 0.611 2.54
60 4.55 169.6 232.6 0.8 2.54 60 2.62 109 178 1 2.54 60 1.65 73.5
137.8 2 2.54 60 0.400 19.7 50 2.54 2.54 60 0.254 12.6 33.8 2.54 3
60 0.189 9.5 26 2.54 3.5 60 0.146 7.3 20.5 2.54 4 60 0.118 5.9
16.7
EXAMPLE 3
[0098] A 60 cm long source is covered with an aperture plate and
heater assembly having 60 apertures with their centers spaced 0.9
cm apart, starting 2.55 cm from the end of source. The apertures
are rectangular slots 0.0125 cm wide.times.0.5 cm long. The source
body is rectangular in cross section, having width b and depth a,
where the depth a denotes the distance from the aperture plate to
the surface of the material loaded in the source. The deposition
rate is 1 A/s at an aperture-substrate distance of 12 cm, and the
material being vaporized is aluminum tris-quinolate (AlQ) (having a
molecular mass of 459, a molecular diameter of 1 nm, and a vapor
pressure curve as given in FIG. 7). Values for a and b are shown in
Table III. The conductance ratio is calculated for the required
operating pressure to produce the specified rate and is shown in
Table III. The pressure nonuniformity is calculated as described
above. Results are shown in Table III for the case where there are
no leaks (other than the apertures) and for the case where there is
a leak at one end of the source, said leak having a conductance
equal to the total conductance of all the apertures.
3TABLE III Source Width, b (cm) Source depth, a (cm) Source Length
(cm) 23 Conductance Ratio : C apertures C sourcebody ( C A / C B )
24 Pressure nonuniformity 100 ( P max - P min ) ( P max + P min ) /
2 Pressure nonuniformity with leak C.sub.leak = C.sub.A 0.611 2.54
60 5.71 200.1 257.7 0.8 2.54 60 3.67 143.7 210.3 1 2.54 60 2.56
107.5 175.9 2 2.54 60 0.863 40.9 90.5 2.54 2.54 60 0.589 28.5 68.1
2.54 3 60 0.457 22.3 55.7 2.54 3.5 60 0.361 17.8 45.9 2.54 4 60
0.296 14.7 38.7
EXAMPLE 4
[0099] A 60 cm long source is covered with an aperture plate and
heater assembly having 60 apertures with their centers spaced 0.9
cm apart, starting 2.55 cm from the end of source. The apertures
are circular holes having a 0.04 cm diameter. The source body is
rectangular in cross section, having width b and depth a, where the
depth a denotes the distance from the aperture plate to the surface
of the material loaded in the source. The deposition rate is 50 A/s
at an aperture-substrate distance of 12 cm, and the material being
vaporized is aluminum tris-quinolate (AlQ) (having a molecular mass
of 459, a molecular diameter of 1 nm, and a vapor pressure curve as
given in FIG. 7). Values for a and b are shown in Table IV. The
conductance ratio is calculated for the required operating pressure
to produce the specified rate and is shown in Table IV. The
pressure nonuniformity is calculated as described above. Results
are shown in Table IV for the case where there are no leaks (other
than the apertures) and for the case where there is a leak at one
end of the source, said leak having a conductance equal to the
total conductance of all the apertures.
4TABLE IV Source Width, b (cm) Source depth, a (cm) Source Length
(cm) 25 Conductance Ratio : C apertures C sourcebody ( C A / C B )
26 Pressure nonuniformity 100 ( P max - P min ) ( P max + P min ) /
2 Pressure nonuniformity with leak C.sub.leak = C.sub.A 0.611 2.54
60 0.427 20.9 52.7 0.8 2.54 60 0.219 10.9 29.6 1 2.54 60 0.126 6.3
17.8 2 2.54 60 0.0254 1.3 3.8 2.54 2.54 60 0.0156 0.8 2.3 2.54 3 60
0.0114 0.6 1.7 2.54 3.5 60 0.00869 0.44 1.3 2.54 4 60 0.00696 0.35
1
EXAMPLE 5
[0100] A 30 cm long source is covered with an aperture plate and
heater assembly having 30 apertures with their centers spaced 0.9
cm apart starting 0.96 cm from the end of source. The apertures are
rectangular slots 0.0125 cm wide.times.0.5 cm long. The source body
is rectangular in cross section, having width b and depth a, where
the depth a denotes the distance from the aperture plate to the
surface of the material loaded in the source. The deposition rate
is 10 A/s at an aperture-substrate distance of 12 cm, and the
material being vaporized is aluminum tris-quinolate (AlQ) (having a
molecular mass of 459, a molecular diameter of 1 nm, and a vapor
pressure curve as given in FIG. 7). Values for a and b are shown in
Table V. The conductance ratio is calculated for the required
operating pressure to produce the specified rate and is shown in
Table V. The pressure nonuniformity is calculated as described
above. Results are shown in Table V for the case where there are no
leaks (other than the apertures) and for the case where there is a
leak at one end of the source, said leak having a conductance equal
to the total conductance of all the apertures.
5TABLE V Source Width, b (cm) Source depth, a (cm) Source Length
(cm) 27 Conductance Ratio : C apertures C sourcebody ( C A / C B )
28 Pressure nonuniformity 100 ( P max - P min ) ( P max + P min ) /
2 Pressure nonuniformity with leak C.sub.leak = C.sub.A 0.611 2.54
30 1.39 63.9 124.6 0.8 2.54 30 0.810 38.9 86.7 1 2.54 30 0.514 25.3
61.4 2 2.54 30 0.128 6.5 18.2 2.54 2.54 30 0.0817 4.2 11.9 2.54 3
30 0.0611 3.1 9 2.54 3.5 30 0.0472 2.4 7 2.54 4 30 0.0381 2 5.7
[0101] Experimental data are provided in FIG. 9 for comparison with
the example calculations, specifically FIG. 8, where pressure
nonuniformity vs. deposition rate data is plotted for different
cross-sectional areas with and without end leaks from Examples 1-5.
In the experiment, a source of 50 cm length was filled with
material (AlQ) such that its body cross sectional area was roughly
10 cm.sup.2. An aperture plate and heater assembly with a series of
49 rectangular apertures, 0.0125 cm wide by 0.5 cm long was used
with this source body. Near the center of the source, the centers
of the apertures were spaced roughly 1 cm apart. The apertures near
the edges were spaced more closely together to compensate for edge
effects in the uniformity profile (as disclosed by Freeman et al,
U.S. application Ser. No. 10/093,739, filed Mar. 8, 2002. This
particular construction had zones near the ends where the operating
temperatures were significantly cooler than the central region (by
roughly 25 C).
[0102] On initial warm-up of the source, material condensing at the
source ends produced a significant mass flow that simulated an
effective leak at both ends of the source. An array of quartz
crystal monitors placed 10 cm above the source along its length
recorded the rate profile, and thus the uniformity profile of
material emanating from the source. Nonuniformity was thus measured
as a function of source operating rate (i.e., deposition rate in
A/s). An initial set of data was taken during the first warm-up of
the source. After hours of operation, material accumulated at the
ends of the source to such an extent that it was closer to the
heater assembly, and therefore more effectively heated, thereby
reducing the effective leak arising from the condensation effects.
At this point the source operation was stable, and nonuniformity
data were acquired as a function of operating rate. As can be seen
from FIG. 8, the data are consistent with the calculations
presented in the examples.
[0103] The above examples illustrate the dependence of source
performance (% Pressure Nonuniformity) on source body cross
sectional area and operating rate of the source. A graph
summarizing the calculated pressure drops from all the examples as
a function of the cross-sectional area of the source body is shown
in FIG. 10. Prior to this invention, those skilled in the art would
construct sources of varying body dimensions and aperture sizes and
then empirically select the appropriate geometry based on desired
uniformity and the required rate range. While the rate dependence
of uniformity would not be entirely obvious to those skilled in the
art, the general procedure of source optimization would accommodate
this phenomenon by sizing the source body and apertures for the
intended deposition rates and intended substrate sizes to be
coated.
[0104] The examples presented above, when analyzed in terms of a
single parameter--the conductance ratio, illustrate that robust
source operation can be achieved by design according to the present
invention. Shown in FIG. 11 are all the data for pressure
nonuniformity as a function of conductance ratio for all the
examples presented above. As can be seen from the graph, the
previously demonstrated rate dependence and geometrical dependence
of performance (see Tables I-V and FIG. 10) are accounted for
entirely by the conductance ratio. As long as the conductance ratio
is below a specified value, the performance will equal or exceed
the specified uniformity requirements. In the examples presented
above, the mechanism by which uniformity worsens at lower rates is
the loss of viscous flow contribution to the conductance along the
source body. Specifying source body and aperture geometries to work
in this low-rate (molecular flow) limit will produce sources that
perform over a much wider range of rates than can be achieved by
constructing a source that just meets uniformity requirements at
high rates and then operating the source at lower rates.
[0105] By this approach, a single source body and aperture
construction can be devised to perform at all rates above a
specified minimum rate and below some practical limit determined by
the operating temperatures at high rates (and thus high pressures)
as dictated by the vapor pressure curve for the material being
deposited.
[0106] In accordance with the present invention, a source can be
designed to work in the low-pressure limit of molecular flow by
appropriate control of conductance ratio through aperture size and
source body cross sectional dimensions. This novel approach avoids
the need for multiple sources for differing ranges of required
deposition rate.
[0107] Also, in accordance with the present invention, a source can
also be designed to have the appropriate conductance ratio by
providing a suitable pressure of inert gas, such as argon, in the
source body. This approach would be advantageous in cases where
there are significant restrictions on the maximum source body
dimensions or significant limitations on the ability to produce
sufficiently small apertures. This approach boosts the source
operating pressure, thereby attaining transition or viscous flow
along the source axis. Because this pressure is achieved by means
other than increasing the vapor pressure (or partial pressure Pm)
of the material to be deposited (i.e. by heating the material
further), it has additional advantages for materials that are
highly susceptible to thermal degradation. Geometrical restrictions
and thermal degradation notwithstanding, in the case where a source
operates well (i.e., with satisfactorily low non-uniformity) at
high rates but not at low rates, an inert gas can be introduced to
the source body to maintain some viscous flow in the source body
down to the lowest rates required. This approach avoids the need to
construct a different source with smaller apertures to limit the
rate resulting from the need to operate at high vapor pressures to
achieve suitable uniformity.
[0108] A flow chart summarizing the calculation of conductances and
pressure distribution for a given deposition rate and geometry (as
described in detail earlier and as used to generate the examples
presented above) is shown in FIG. 12. First, a source throughput
per unit length required for a desired deposition rate 222 at a
desired source-substrate spacing 220 and a given total aperture
area 224 is calculated 200. Next, using the individual aperture
area 224 and aperture length 226, an initial guess of vapor
temperature 218, and the molecular mass of the material 228, the
internal pressure required to produce the throughput 200 is
calculated 202. Using a molecular diameter 230, the guess of the
temperature 218, and the internal pressure 202 are used to
calculate a mean free path 204. The aperture dimensions 232 and the
source body dimensions 234 are used in combination with the mean
free path 204 to calculate a Knudsen number 206. The dimensions 232
and source body dimensions 234 as well as the molecular mass 228
and molecular diameter 230 are used with the appropriate
conductance formula chosen based on the calculated Knudsen number
to determine the respective aperture and source body conductances
and a ladder network model is formed 208. Pressure distribution in
the ladder network is calculated 210. Vapor temperatures are
estimated 212 from a vapor pressure curve. The estimate 212 of the
temperature is compared 216 to the initial guess 218. If the
estimate 212 is significantly different (e.g. more than 2% of
absolute temperature) from the initial guess 218, the guess is
updated to equal the estimate and the calculations are repeated
until the guess and the estimate are in agreement. A final value
for pressure uniformity, average pressure, and vapor temperature is
then output 214. For given material, deposition geometry, and
deposition rate, this process is repeated for a variety of aperture
and source body dimensions to produce graphs like those shown in
FIGS. 10 and 11. A suitable design is then chosen from these
graphs.
[0109] The invention has been described in detail with particular
reference to certain preferred embodiments thereof, but it will be
understood that variations and modifications can be effected within
the spirit and scope of the invention.
Parts List
[0110] 1 source body
[0111] 2 material
[0112] 3 cover
[0113] 4 aperture
[0114] 5 end contacts
[0115] 6 baffle
[0116] 10 source body
[0117] 11 material
[0118] 12 cover
[0119] 13 aperture
[0120] 14 end contacts
[0121] 15 baffle
[0122] 16 slit
[0123] 30 material
[0124] 31 source body
[0125] 32 aperture plate
[0126] 33 aperture
[0127] 34 baffle
[0128] 35 source--substrate distance
[0129] 40 substrate
[0130] 41 carriage
[0131] 42 standoffs
[0132] 50 radiation shield
[0133] 100 vapor deposition station
[0134] 101 housing
[0135] 102 chamber
[0136] 110 substrate
[0137] 111 mask frame
[0138] 114 deposition source
[0139] 115 insulating source support
[0140] 116 electrical leads
[0141] 117 electrical leads
[0142] 120 feedthroughs
[0143] 121 feedthroughs
[0144] 130 material
[0145] 131 deposition zone
[0146] 132 aperture
[0147] 140 glide shoe
[0148] 142 lead screw follower
[0149] 145 lead screw
[0150] 147 glide rail
[0151] 149 shaft seal
[0152] 150 lead screw shaft
[0153] 151 motor
[0154] 200-218 steps
[0155] 220 source substrate spacing
[0156] 22 deposition rate
[0157] 224 aperture area
[0158] 226 aperture length
[0159] 228 molecular mass of material
[0160] 230 molecular diameter
[0161] 232 aperture dimensions
[0162] 234 source body dimensions
* * * * *